From the file marked ‘Why journalists shouldn’t do statistics’ we have a new front page entry from the Daily Mail.

As per standard Ministry policy, I’m not going to link to the Mail’s article but I will quote from it as I tear it into little tiny kibbles – and as your starter for ten we’ll be tackling the headline claim that the NHS spends £1 million a week on ‘repeat abortions’.

The first thing to not is that the article does not provide a clear  source for this claim straight away. All we get at the beginning of the article is a series of general claims and statements, as follows:

The Health Service is spending around £1million a week providing repeat abortions.

Critics said figures revealed yesterday show thousands of women are using the procedure as a form of contraception.

It is not unknown for some women to have seven, eight or even nine terminations in their lifetime.

According to the statistics, single or unmarried women account for five out of every six repeat terminations. Around a third of all abortions carried out in England and Wales are repeats.

The figures will fuel the debate on whether abortions, which cost the NHS up to £1,000 each, are being sanctioned as more of a lifestyle choice than a medical requirement.

In one London borough, half of all abortions are requested by women who have already aborted at least one foetus.

In 2010, the latest year for which figures are available, some 189,000 abortions took place. Of these, more than 64,000 terminations were on women who had already aborted a foetus in the past.

More than 50,000 of them were single or living with a partner; while around 9,500 were married – with the marital status unknown for the rest.

Critics of the current legislation said terminations have become just another form of contraception because of the lack of independent counselling at clinics which leads to women making the choice to have an abortion rather than continue with their pregnancy.

*There a photo of Diane Abbott on the article, which is captioned ‘Revelations: The figures were revealed in a parliamentary question submitted by Labour health spokesman Diane Abbott’ but on examination of the parliamentary record it appears that the only figures obtained by Abbott were figures for repeat abortions by age and by age and marital status. Her question had nothing whatsoever to do with costs.

The ‘critics’ referred to in the article by name are –

1. Josephine Quinataville, a former journalist, ‘crisis pregnancy counsellor’ with LIFE and (reportedly) multi-millionairess who gets treated by the right-wing press as an ‘expert’ on reproductive ethics on the back of her role as director of a self-founded and financed organisation, Comment on Reproduction Ethics (CORE) despite the fact that she appears to have no discernable qualifications in the field of bioethics, and

2. Stuart Cowie, who’s billed as ‘a spokesman for pressure group LIFE’ although he’s actually their Head of Education and sports the following biographical information on LIFE’s website:

A graduate in International Development, Stuart brings a wide range of experience, from other charities and from the private sector, to his role as Head of department. He is a family man, with great enthusiasm for the trombone, which he plays in a funk band.

All of which suggests that the ‘critics’ mentioned in the article, or rather their parent organisations – Quintaville is billed is a member of the Pro-Life Alliance of which LIFE is prominent member – may well be the source of some of these claims.

When we do eventually get to some information on the source of these claims, what we’re told is that:

Figures late last year showed that each abortion costs the NHS an average of £680. It means around £850,000 is spent on repeat abortions every single week in England and Wales.

The figures do not include Scotland and Northern Ireland, which means the full UK cost will be around £1million a week.

Bullshit – elective abortions are still illegal in Northern Ireland and the figures for Scotland show that in 2010 there were 12, 836 abortions carried out in Scotland of which 28.4% (3645) were ‘repeat abortions’. Based on the Daily Mail’s figures, the average unit cost of an abortion in Scotland works out at a whopping £2,149.

The claim that it costs the NHS in England and Wales £850,000 a week to provide ‘repeat abortions’ equates, of course, to a cost of £44.2 million a year, and for 2010, the year from which the Mail is taking its figures, the number of women who had an abortion having already had at least one previous abortion was 64,303 (source: NHS abortion statistics 2010). Dividing the two figures give an average unit cost of just over £687 and further evidence of the Daily Mail playing fast and loose with the figures. To get their main estimate they’ve rounded up the number of abortions to 65,000 which creates a rounding error of £476,000 in the annual figure.

Even without chucking an extra 700 abortions into the figures, the Mail’s numbers are complete bullshit, particular the claim that abortions can cost the NHS up to £1000 each. This is true only of non-elective abortions, which are considerably more expensive that elective procedures, with tariffs for surgical abortion ranging from £742 (vacuum aspiration at less than 9 weeks) to £1,543 for a late term D&E, but these are for non-elective procedures, which are undertaken following a miscarriage or where the foetus has died in-utero. The highest NHS tariff for an elective abortion in 2010 was only £662 and with 92% of abortions taking place before 12 weeks gestation, the overwhleming of majority of abortions would have been carried out at NHS tariffs of between £470 and £560.

This is an issue I looked at in some detail last year and back then I found that even with government incorrectly including the costs on non-elective procedures in theeir figures they still managed to come up with an average cost per abortion of £680, which would give an annual cost of £43.7 million for ‘repeat’ abortions if you don’t try and fiddle the figures, and when I re-did the figure using the government’s own calculation methods, but excluding non-elective procedures, the average cost per abortion came down to just £580 per abortion, giving an annual cost for ‘repeat abortions’ of just £37.3 million.

Even that lower figure may be an overestimate as I used a base cost of £494, extrapolated from figures given in BPAS’s accounts, only to discover some months later that BPAS actually charges the NHS around £420 per abortion. Based on that last figure, the average cost per abortion comes down to around £515 with an annual cost for ‘repeat abortions’ of £33-34 million a year.

On a claimed figure of £44.2 million a year for England and Wales, the Daily Mail’s figures are out by anywhere from £6-10 million a year.

Factoring in Scotland adds either £2.11 million (at £580) or £1.87 million (at £515) to our estimates giving us a UK total of either £35 or £39 million, depending on which base cost we use, compared to the Daily Mail’s £52 million headline figure – the Mail is out by £13-£16 million.

The second key theme in the Mail’s article is the perennial favourite which claims that women are using abortion as a form of contraception and as supporting ‘evidence’ we’re offered only an example of ‘proof by anecdote’, a callout box that relates the story of a young woman who claims to have had four abortions by the age of 16.

Claims such as this are impossible to verify from official statistics as not breakdown data is given for abortions by age and number of abortions where there are fewer than 10 women in a particular group but in regards to the statement that its “not unknown for some women to have seven, eight or even nine terminations in their lifetime” the figures for 2010 show that 85 women had had seven or more previous abortions, with a further 69 having had six previous abortions.

This, of course, tells nothing at all about the circumstances which led to these women having so many abortions – this is an issue that, like so many others, remains woefully underresearched in the UK due to such research being implicitly discouraged by the outdated medico-legal framework under which abortion services are required to operate in the UK. That said, what I have been told in the past by NHS sexual health workers is that amongst the minority of women who do present for a fifth, sixth or seventh abortion, its not uncommon to find street prostitutes – unprotected sex often pays better – women with learning difficulties, women with substance or alcohol abuse problems, victims of domestic abuse/violence and, sadly, women with two or more of these issues to deal with.

As is invariably the case, the only context we’re given in which to judge the claim that women are using abortion as form of contraception is an anecdote about a women who, on the face of it, appears to have resorted to abortion for just that purpose. However, anecdotes are not the only the way we can judge such claims, providing we’re willing to crunch a few numbers.

Let’s start with a very basic statistic – in 2010, the estimated female population of England and Wales was 27,155,000 of which 12,387,000 were aged between 16 and 49 years of age.

Based on NHS estimates, at any given time around 57-58% of women aged 16-49 use some form of non-surgical contraception; the contraceptive pill and male condoms account for around 25% each with IUDs and other contraceptive devices making up the other 7-8%.

So, in 2010, there were just over 3 million ahed 16-49  ‘on the pill’ and around the same number relying on male condoms in order to avoid falling pregnant.

Neither of these contraceptive methods is, of course, 100% effective and so far as the failure rate for oral contraceptives and condoms is concerned the figures that most people will have encountered are those for ‘perfect use’. If used correctly, all the time, the failure rate for oral contraceptives, i.e. the percentage of women who fall pregnant while taking the pill, is around 0.3%, while for condoms its around 2%.

Those are the ideal figures, but what about what happens out in the real world where people aren’t quite so perfect?

While we don’t have figures covering every possible circumstance in people are less than perfect in their use of contraception what we do have are figures for the failure rates for the ‘typical’ use of different types of contraception within the first year of usage, and these tell a bit of different story. In typical usage, the failure rate for oral contraceptives is 8%, while for condoms its 15% – and given some of the propaganda surrounding condom use I should stress that this difference between perfect use and typical use is down to user ‘error’ and the typical failure rate for ‘natural’ family planning is around 25%.

Unfortunately we don’t have longitudinal figures for failure rates in typical use, so its not clear whether or to what extent ‘practice makes perfect’ but, for illustrative purposes, we’re going to assume the same failure rates across the board, regardless of length of usage, and when we map these failure rates on to the figures for contraceptive use we get the following estimates for the maximum annual number of unintentional pregnancies that result from contraceptive failures.

Oral contraceptives – 247,700

Male condoms – 464,500

So that’s a maximum of 713,000 or so unintended pregnancies for 2010 out of…

Okay, next calculation… for 2010 we have just over 723,000 live births and just under 190,000 abortions, and then there are miscarriages which account for around 15-20% of recognised pregnancies – although as many as 50% of all conceptions end in an early spontaneous abortion which go unrecognised these are not relevant to the figures we’re working with which deal only with pregnancies that are recognised. Taking a conservative estimate of the number of miscarriages (15%) gives us a figure of 161,000 and a total number of pregnancies for the year of around 1.074 million.

On paper that’s a lot of ‘accidents’ as it suggest that as many as two thirds of pregnancies are unplanned or, at least, unexpected – remember, at least some of these pregnancies will occur in couples who do plan to have children but who haven’t started deliberately trying for a baby at the point when conception occurs.

What this gives us is a breakdown of outcomes – of these pregnancies, around 67% result in a live birth, 18% in an abortion and 15% in a miscarriage, and if we apply those percentages to our figures for conceptions due to contraceptive failures we get an estimate of 43,700 abortions resulting from oral contraceptive failures and 82,000 from condom failures giving a total of 125,700 abortions that are attributable to contraceptive failures.

That’s two thirds of the total number of resident abortions in England and Wales in 2010.

Okay, so in terms of accuracy, we have a game of swings and roundabouts here. It’s entirely possible, if not likely, that contraceptive failures account for lower number of pregnancies than our, admittedly, top end estimate and people are less likely to make mistakes the longer they use a particular contraceoptive method, although its also possible that any gains we accrue from experience could be offset by losses due to complacency.

We’re also assuming that conceptions arising from a contraceptive failure are no more, or less, likely to result in an abortion than other conceptions when that may well not be true – I think most people would assume that a pregnancy resulting from a contraceptive failure is more likely to result in an abortion simply because active steps were being taken to try and prevent the pregnancy in the first place.

The best we can say, therefore, is that we have some ballpark figures to work with that suggest that anything up to two-thirds of the pregnancies that end in abortion may be the consequence of a contraceptive failure, all of which blows a vey large hole in the claim that abortion is being used as a primary mean of contraception.

This is, of course, all before we start to consider other reasons that come into play when women decide to have an abortion.

Sadly we don’t have research from the UK to work with, but some insights can be gleaned from research conducted in the US which indicates that pregnancies resulting from rape may account for around 1% of abortions – that’s around 1900 abortions a year that don’t fit the contraception claim for starters.

Incest is relatively rare (less than 0.5%) but 2% of the women in the US study cited domestic violence/abuse as a reason why they had an abortion, 12% cited personal health problems, 11% believed that the break-up of their marriage/relationship was imminent and 38% had fallen pregnant after they’d decided that they didn’t want any more children – of the 64,000 women in England and Wales who had a ‘repeat abortion’ in 2010, more than 11,000 were over the age of 35.

Overall, half of all women had an abortion in 2010 already had at least one child and amongst the over 35s, 85% had already had kids.

However you look at it, women who do genuinely use abortion as a primary means of contraception are, contrary to the anti-abortion lobby’s claims, very much in the minority while most women rely on abortion as a last resort when contraception fails or when they find themselves in an extremely difficult and, in some cases, dangerous situations.

The Daily Mail, on the other hand, is full of shit – as usual.